It occurred to me that there might be some doubting Thomases about this double the length and get eight times the deflection thing. So I set up some tube and a weight and measured the difference in deflection between a foot of overhang and two feet of overhang. I used a former for my bender as a weight,

Fig. 3
A foot of tube with no weight

Fig. 4
A foot of tube with the weight

Fig. 5
Two feet of tube with no weight

Fig. 6
Two feet of tube with the weight

As you can see in Fig. 6 the tube is clearly bending way more than it is in Fig. 4. The data in Fig. 7 are the actual measurements I took, and if you multiply the deflection of the one foot length of tube by 8 you get 40mm. Allowing that I did all this with a tape measure and there may very well have been an error of .25mm in any of the first set of measurements which makes it all pretty close to the expected eight times greater deflection.

So if you take the moment of inertia figure from the calculator on the previous page, what does that mean? Well basically if the tube is 1" OD .120" wall then it's moment of inertia would have been 0.033, for 1.25" OD .097" wall tube it would be 0.059. What this boils down to is that if you have a load on a piece of the 1" OD tube that deflects it by 10mm, and you want to know how far the same load would deflect the 1.25" OD tube then what you do is multiply the 10mm by the moment of inertia for the 1" OD to get 0.33. Then you divide that by the moment of inertia for the 1.25" OD tube, which gives you 5.59 mm (0.33 ÷ 0.059) of deflection. That's 44 % less deflection for the same load with a penalty of just over 1 oz weight per linear foot of tube.

If you poke those numbers into the calculator you'll also see that the cross sectional areas are pretty similar too (which is obvious as the weight per linear foot is very similar) so the tensile strength is very similar for two different sizes of tube that are made of the same steel. As I said earlier, you have to watch using too thin a wall so obviously there is a limit where using larger OD's and thinner walls starts to become increasingly dangerous. Aside from the danger of denting the tube, there is also the problem that thinwall structures tend to fail more catastrophically than thicker walled structures. Absolute minimum wall thickness for road use is probably 2mm or 0.079" (14 gauge), I certainly wouldn't go any thinner than that, and for chop frames I stick to a wall thickness of 0.097" (2.5mm) for the rail and use 0.120" (3mm) for the top tube and center post. If I'm building a frame with 1" OD tubes (usually because it has a small motor and I don't want the frame to "overpower" the motor visually) then I mostly use 0.120" wall for the rails and 1.5" OD 0.120" wall for the top tube.

Those sizes are fairly traditional, and there isn't much point in trying to re invent the wheel, but if you're trying to do something a little out of the ordinary then all the stuff I've been wittering on about might be of some use. There is another consideration to using thinner walled tube and that is the bender you are planning on using. If you really must build a frame with super thin walls then you need to be using a good bender, one of the push through type you see in most tool supermarkets isn't going to bend the tube without kinking the wall, ideally you want a mandrel bender of the type that pulls a support through the inside of the tube as it bends it.

So there you go, you should now have some chance of making an informed decision about tube sizes or at least sorting through some of the BS that you get fed on the subject.